Prostate cancer is the most commonly occurring cancer in men of African descent in the U.S. Even though African American men have the highest prostate cancer rate in the world, there is relatively little information available about the etiologic factors that may explain these rates. Despite the knowledge that prostate cancer occurs at high rates in men of African descent in the Americas, very little data is available regarding the epidemiology of prostate cancer in native African men, even though prostate cancer seems to be prevalent in that population. We hypothesize that a better understanding of prostate cancer risk in African men could elucidate factors affecting the high incidence of prostate cancer in African American men. The objective of the present study is to examine the role of genes that regulate the disposition of testosterone in prostate cancer etiology in men of African descent, and to evaluate whether these genes explain, in part, ethnic differences in prostate cancer rates. These genes include the cytochromes P450 CYP3A4, CYP3A5, CYP3A43, and CYP19, 5alpha-reductase II (SRD5A2), androgen receptor (AR), and the hydroxy-steroid dehydrogenases type II 3beta (HSD3B) and type 17beta (HSD17B). These genes are involved in testosterone metabolism, and thus focus our hypotheses on a relevant, defined set of metabolic pathways. Three specific aims are proposed here. In Specific Aim 1, we will use publicly available databases to evaluate genetic variation at candidate androgen metabolism genes in African American and Senegalese population, including SNP characterization, identification of haplotype tag SNPs (htSNPs) and haplotypes, as well as characterization of linkage disequilibrium blocks. In Specific Aim 2. we will characterize population genetic structure in African American and Senegalese men using genomic markers, and use this information to assess the potential for population stratification in African and African American populations. In Specific Aim 3, we will assess the relationship of genetic variability in case-control analyses of prostate cancer in Africans and African Americans using haplotypes and multivariate genotypes. In order to address these aims in this competing continuation proposal, we will build on our existing epidemiological participant accrual and data management infrastructure that has been developed under previously funded grants (R01-CA85074 to Dr. Rebbeck and R03-CA103359 to Dr. Gueye). Using this previous funding, we have developed comprehensive and comparable protocols for the identification of incident prostate cancer cases and controls in Philadelphia and Dakar. At both locations, risk factor information is obtained from a questionnaire interview, a biosample containing DNA is collected using a non-invasive cheek swab method, and pathology information is collected using a standardized medical record abstraction approach. The resources and data generated through this proposal will be used to determine the role of genes and other risk factors in prostate cancer etiology. An understanding of the complex interplay of genetic variability at multiple loci as well as other factors may improve our understanding of ethnic differences in prostate cancer etiology and risk prediction, and in particular help us to understand why African American men have the highest rates of prostate cancer incidence in the world.